OpenBSD src changes summary for 2016-09-10 ==========================================
distrib/miniroot etc/ksh.kshrc gnu/usr.bin/binutils-2.17 lib/libc regress/usr.sbin sbin/bioctl sbin/mount sbin/mount_ffs sbin/ping sbin/ping6 sys/arch/amd64/amd64 sys/arch/amd64/include sys/arch/amd64/stand/boot sys/arch/amd64/stand/cdboot sys/arch/amd64/stand/efiboot sys/arch/amd64/stand/libsa sys/arch/amd64/stand/pxeboot sys/arch/armv7/imx sys/arch/armv7/sunxi sys/arch/i386/i386 sys/arch/i386/stand/boot sys/arch/i386/stand/cdboot sys/arch/i386/stand/libsa sys/arch/i386/stand/pxeboot sys/arch/powerpc/ddb sys/arch/sparc64/sparc64 sys/arch/sparc64/stand/ofwboot sys/dev sys/dev/pci sys/kern sys/lib/libsa sys/net sys/sys sys/ufs/ffs sys/ufs/ufs usr.bin/signify usr.sbin/iscsid usr.sbin/kvm_mkdb usr.sbin/syspatch == distrib =========================================================== 01/09 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/distrib miniroot ~ install.sub > Fix detection if default route was set by dhclient in v4_defroute(). > Route labels are now enclosed in double quotes (see r1.107 of > src/sbin/route/show.c). > OK claudio, krw > no objections halex (rpe@) ~ install.sub > Ensure that install_files() returns successful if the _tmpsrc > directory does not exist. Otherwise, INSTALL_METHOD is not set > and this information is not sent to the cgi server. > OK halex (rpe@) ~ install.sub > Ensure that previous installer choices are available during upgrades > by moving waitcgiinfo() from install_http() to before install_sets() > in do_upgrade(). > OK halex (rpe@) ~ install.sub > Ensure that a previous cdrom install method is passed through > to install_sets() and that it is presented as default answer. > - fix pattern in waitcgiinfo() to match method=cd0. > - use a separate variable _im to set INSTALL_METHOD > resp might get overwritten in install_cdrom() > OK halex (rpe@) == etc =============================================================== 02/09 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/etc ksh.kshrc ~ ksh.kshrc > Identation > OK halex, krw (rpe@) ~ ksh.kshrc > Remove remnants of rlogin, it's long gone (2002). > OK halex (rpe@) ~ ksh.kshrc > Simplify setting the USER and UID variables. > While here, convert `` to $(). > OK halex (rpe@) ~ ksh.kshrc > Unbreak ksh.kshrc by using a MI way of finding out the console device name. > Found by landry > OK deraadt (rpe@) == gnu =============================================================== 03/09 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/gnu usr.bin/binutils-2.17 ~ ld/emultempl/elf32.em > use an unused attribute on a function argument to avoid -Wself-assign > with clang > ok guenther@ (jsg@) ~ binutils/readelf.c ~ gas/symbols.c ~ gas/config/atof-ieee.c ~ gas/config/tc-i386.c ~ opcodes/i386-dis.c > Bring in FreeBSD changes to prevent clang triggering > -Wstring-plus-int -Wempty-body and -Wshift-negative-value warnings > as binutils is built with -Werror. > ok guenther@ millert@ (jsg@) == lib =============================================================== 04/09 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/lib libc ~ arch/amd64/sys/brk.S ~ arch/amd64/sys/sbrk.S > Clang ignores a .weak directive before a function is actually defined. So > move it from before ENTRY() to after END(). Keeps brk(2) and sbrk(2) weak > when comping libc with clang. > ok guenther@ (kettenis@) == regress =========================================================== 05/09 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/regress usr.sbin ~ ldapd/Makefile > Improve Makefile by using ${.OBJDIR} and stop creating files in > ${.CURDIR}. Don't start the server in .BEGIN - make obj should only > create the symlink. Add CLEANFILES for proper cleaning. (landry@) ~ ldapd/run-tests.pl > Cleanup commented lines, and only print the DN of entries (landry@) == sbin ============================================================== 06/09 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/sbin bioctl ~ bioctl.c > Clean up some softraid crypto code - rename struct sr_crypto_kdf_pbkdf2 to > sr_crypto_pbkdf (since it is useable for more than just pkcs5_pbkdf2) and > embed a struct sr_crypto_genkdf within it, rather than redeclaring the same > fields. Rename SR_CRYPTOKDFT_PBKDF2 to SR_CRYPTOKDFT_PCKS5_PBKDF2 and add > SR_CRYPTOKDFT_BCRYPT_PBKDF for upcoming changes. (jsing@) ~ bioctl.c > Ugh, too many initialisms... (jsing@) ~ bioctl.c > Teach bioctl derive_key() how to handle bcrypt_pbkdf. > Part of a diff from djm@ (jsing@) mount ~ mntopts.h ~ mount.8 ~ mount.c > Add a noperm mount flag for FFS to be used for building release sets > without root privileges. This is only the kernel/mount flag; additional > work in the build Makefile's will be necessary such that the files in > $DESTDIR are created with correct permissions. > tedu couldn't find anything wrong with it in a quick review > idea & ok deraadt (natano@) ~ mount.8 > sort the options list; also, the full stop in (FFS only.) wasn;t > doing much, so i zapped it; (jmc@) mount_ffs ~ mount_ffs.c > Add a noperm mount flag for FFS to be used for building release sets > without root privileges. This is only the kernel/mount flag; additional > work in the build Makefile's will be necessary such that the files in > $DESTDIR are created with correct permissions. > tedu couldn't find anything wrong with it in a quick review > idea & ok deraadt (natano@) ping ~ ping.c > reorder function declaration to reduce diff between ping and ping6; no > object change (florian@) ~ ping.c > packet is actually u_char, pass it around as one (florian@) ~ ping.c > reorder main variables; no obj change (florian@) ~ ping.c > reorder main variables, reduces diff (florian@) ~ ping.c > reorder main variables, no object change, reduces diff to ping6 (florian@) ~ ping.c > The getopt while loop is very long, put the body in {}. > Removes a "Huh, that's weird" diff to ping6. No object change. (florian@) ping6 ~ ping6.c > reorder function declaration to reduce diff between ping and ping6; no > object change (florian@) ~ ping6.c > reorder main variables; no obj change (florian@) ~ ping6.c > reorder main variables, reduces diff (florian@) ~ ping6.c > move IPv6 specific structs up; no obj change (florian@) ~ ping6.c > according to netinet6/in6.h IPV6_MULTICAST_LOOP setsockopt takes a > u_char argument (florian@) == sys =============================================================== 07/09 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/sys arch/amd64/amd64 ~ db_trace.c > take it one step further and bring the message inline with arm/sparc64 > pointed out by guenther@ (jasper@) ~ vmm.c > Enable VMM debug and add a few new controls (mlarkin@) ~ vmm.c > Skip printing a debug message when an external interrupt is received > (mlarkin@) arch/amd64/include ~ specialreg.h > Enable VMM debug and add a few new controls (mlarkin@) arch/amd64/stand/boot ~ Makefile > Rename libsa pbkdf2.c to pkcs5_pbkdf2.c so that we match libutil. (jsing@) arch/amd64/stand/cdboot ~ Makefile > Rename libsa pbkdf2.c to pkcs5_pbkdf2.c so that we match libutil. (jsing@) arch/amd64/stand/efiboot ~ Makefile.common > Rename libsa pbkdf2.c to pkcs5_pbkdf2.c so that we match libutil. (jsing@) arch/amd64/stand/libsa ~ softraid.c > Rename libsa pbkdf2.c to pkcs5_pbkdf2.c so that we match libutil. (jsing@) ~ softraid.c > Clean up some softraid crypto code - rename struct sr_crypto_kdf_pbkdf2 to > sr_crypto_pbkdf (since it is useable for more than just pkcs5_pbkdf2) and > embed a struct sr_crypto_genkdf within it, rather than redeclaring the same > fields. Rename SR_CRYPTOKDFT_PBKDF2 to SR_CRYPTOKDFT_PCKS5_PBKDF2 and add > SR_CRYPTOKDFT_BCRYPT_PBKDF for upcoming changes. (jsing@) arch/amd64/stand/pxeboot ~ Makefile > Rename libsa pbkdf2.c to pkcs5_pbkdf2.c so that we match libutil. (jsing@) arch/armv7/imx ~ imxiomuxc.c > Correct the path to the iomuxc fdt node so pinctrl setup will run. > Noticed when looking into a regression on sabre lite with fec(4) > bisected by matthieu@. This fixes the sabre lite problems with fec and > also recent problems with the i2c rtc on cubox. > ok kettenis@ (jsg@) arch/armv7/sunxi ~ sximmc.c > Moving the host reset after setting the clock in the attach path makes > the host reset succeed on the allwinner,sun5i-r8. > ok kettenis (mglocker@) ~ sxiccmu.c > Use PLL6 as a parent clock for the SDx clocks for frequencies > 400 kHz. > This makes sximmc(4) much faster since we can actually provide the > appropriate > clock rates for the modern SD cards. This uncovered a bug in the code that > calculated the PLL6 output frequencies, which this commit also fixes. > (kettenis@) ~ sximmc.c > Advertise high-speed support. (kettenis@) arch/i386/i386 ~ db_trace.c > take it one step further and bring the message inline with arm/sparc64 > pointed out by guenther@ (jasper@) arch/i386/stand/boot ~ Makefile > Rename libsa pbkdf2.c to pkcs5_pbkdf2.c so that we match libutil. (jsing@) arch/i386/stand/cdboot ~ Makefile > Rename libsa pbkdf2.c to pkcs5_pbkdf2.c so that we match libutil. (jsing@) arch/i386/stand/libsa ~ softraid.c > Rename libsa pbkdf2.c to pkcs5_pbkdf2.c so that we match libutil. (jsing@) ~ softraid.c > Clean up some softraid crypto code - rename struct sr_crypto_kdf_pbkdf2 to > sr_crypto_pbkdf (since it is useable for more than just pkcs5_pbkdf2) and > embed a struct sr_crypto_genkdf within it, rather than redeclaring the same > fields. Rename SR_CRYPTOKDFT_PBKDF2 to SR_CRYPTOKDFT_PCKS5_PBKDF2 and add > SR_CRYPTOKDFT_BCRYPT_PBKDF for upcoming changes. (jsing@) arch/i386/stand/pxeboot ~ Makefile > Rename libsa pbkdf2.c to pkcs5_pbkdf2.c so that we match libutil. (jsing@) arch/powerpc/ddb ~ db_trace.c > take it one step further and bring the message inline with arm/sparc64 > pointed out by guenther@ (jasper@) arch/sparc64/sparc64 ~ db_trace.c > Delete support for 32bit frame backtracing > Convert function definitions from K&R to Standard C > Delete superfluous casts and fix some whitespace > ok jasper@ mpi@ kettenis@ deraadt@ (guenther@) ~ trap.c > If syscall() gets a 32bit trapframe, just kill the process; delete the > other code for 32bit processes, which collapses SYS_syscall and > SYS___syscall > into the same behavior > ok kettenis@ deraadt@ (guenther@) ~ trap.c > Simplify comments to reflect that we don't do 32bit compat processes > ok kettenis@ (guenther@) ~ trap.c > Convert function definitions from K&R to standard-style > ok kettenis@ (guenther@) arch/sparc64/stand/ofwboot ~ Makefile ~ softraid.c > Rename libsa pbkdf2.c to pkcs5_pbkdf2.c so that we match libutil. (jsing@) ~ softraid.c > Clean up some softraid crypto code - rename struct sr_crypto_kdf_pbkdf2 to > sr_crypto_pbkdf (since it is useable for more than just pkcs5_pbkdf2) and > embed a struct sr_crypto_genkdf within it, rather than redeclaring the same > fields. Rename SR_CRYPTOKDFT_PBKDF2 to SR_CRYPTOKDFT_PCKS5_PBKDF2 and add > SR_CRYPTOKDFT_BCRYPT_PBKDF for upcoming changes. (jsing@) dev ~ softraidvar.h > Clean up some softraid crypto code - rename struct sr_crypto_kdf_pbkdf2 to > sr_crypto_pbkdf (since it is useable for more than just pkcs5_pbkdf2) and > embed a struct sr_crypto_genkdf within it, rather than redeclaring the same > fields. Rename SR_CRYPTOKDFT_PBKDF2 to SR_CRYPTOKDFT_PCKS5_PBKDF2 and add > SR_CRYPTOKDFT_BCRYPT_PBKDF for upcoming changes. (jsing@) ~ softraidvar.h > Ugh, too many initialisms... (jsing@) dev/pci ~ if_iwm.c ~ if_iwmreg.h > Move IWM_NVM_* macro definitions from if_iwm.c to if_iwmreg.h. (stsp@) ~ if_iwm.c > Rename iwm_mvm_sta_send_to_fw() to iwm_mvm_sta_cmd(). Get rid of pointless > wrapper functions iwm_mvm_add_sta() and iwm_mvm_update_sta(). (stsp@) ~ if_iwm.c > Fix formatting in iwm such that > break; } > becomes: > break; > } (stsp@) ~ if_iwm.c ~ if_iwmreg.h ~ if_iwmvar.h > iwm(4) inherited the 'MVM' prefix in many of its symbol names from iwlwifi. > Linux iwlwifi implements support for iwn(4) chips in the 'DVM' driver, > and support for iwm(4) chips in the 'MVM' driver. > In OpenBSD the 'MVM' prefix is redundant, so we can remove it to shorten > many lines and reduce noise: s/iwm_mvm_/iwm_/g (stsp@) ~ if_iwm.c > Rename the global nvm_to_read const int array to iwm_nvm_to_read. (stsp@) ~ if_iwm.c > Use consistent indentation for multi-line function prototypes in iwm. > (stsp@) kern ~ vfs_syscalls.c > Add a noperm mount flag for FFS to be used for building release sets > without root privileges. This is only the kernel/mount flag; additional > work in the build Makefile's will be necessary such that the files in > $DESTDIR are created with correct permissions. > tedu couldn't find anything wrong with it in a quick review > idea & ok deraadt (natano@) lib/libsa ~ pbkdf2.c ~ pbkdf2.h > Sync libsa pkcs5_pbkdf2() with libutil. (jsing@) - pbkdf2.c - pbkdf2.h + pkcs5_pbkdf2.c + pkcs5_pbkdf2.h > Rename libsa pbkdf2.c to pkcs5_pbkdf2.c so that we match libutil. (jsing@) + sha2.c + sha2.h > Add sha2 to libsa, taken from libc - needed for bcrypt_pbkdf. (jsing@) + blowfish.c + blowfish.h > Add blowfish to libsa, taken from libc - needed for bcrypt_pbkdf. (jsing@) + bcrypt_pbkdf.c + bcrypt_pbkdf.h > Add bcrypt_pbkdf to libsa, from libutil. This will soon allow the boot > loader to support softraid crypto volumes using bcrypt pbkdf. (jsing@) net ~ bfd.c > Remove a semicolon after an if statement that resulted in > bfd_senddown() always returning before the rest of the function > could run. Found with clang. > ok claudio@ (jsg@) sys ~ mount.h > Add a noperm mount flag for FFS to be used for building release sets > without root privileges. This is only the kernel/mount flag; additional > work in the build Makefile's will be necessary such that the files in > $DESTDIR are created with correct permissions. > tedu couldn't find anything wrong with it in a quick review > idea & ok deraadt (natano@) ufs/ffs ~ ffs_vnops.c > Add a noperm mount flag for FFS to be used for building release sets > without root privileges. This is only the kernel/mount flag; additional > work in the build Makefile's will be necessary such that the files in > $DESTDIR are created with correct permissions. > tedu couldn't find anything wrong with it in a quick review > idea & ok deraadt (natano@) ufs/ufs ~ ufs_lookup.c ~ ufs_vnops.c > Add a noperm mount flag for FFS to be used for building release sets > without root privileges. This is only the kernel/mount flag; additional > work in the build Makefile's will be necessary such that the files in > $DESTDIR are created with correct permissions. > tedu couldn't find anything wrong with it in a quick review > idea & ok deraadt (natano@) == usr.bin =========================================================== 08/09 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/usr.bin signify ~ signify.c ~ zsig.c > spaces found during re-read (deraadt@) == usr.sbin ========================================================== 09/09 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/usr.sbin iscsid ~ control.c > fix an fd leak in an error path (jsg@) kvm_mkdb ~ nlist.c > missing fclose() in an error path (jsg@) syspatch ~ syspatch.8 ~ syspatch.sh > apply_patches should not take any arguments. > Few tweaks while here. (ajacoutot@) ~ syspatch.sh > Drop syspatch_trap, it's simple enough for trap no needing its own > function. > Introduce sp_err() which will get used for error messages. (ajacoutot@) =============================================================================== _______________________________________________ odc mailing list [email protected] http://www.squish.net/mailman/listinfo/odc
